Integrity Monitoring and Prediction Concept and Prototype for Fully Autonomous Vehicle Resilience and Safety

R. Tiwari, T. Stacey, Felix Toran

Peer Reviewed

Abstract:	The Society of Automotive Engineers (SAE) have 5 defined levels of automation within the J3016 standard for vehicular automation within a dynamic environment. To achieve the maximum level of automation, the vehicle must be able to rely on its onboard sensor(s) to navigate in real-traffic environment. Single and/or only augmentation system may not be sufficient to meet safety and integrity of road user. Where data fusion through a multi-sensor, including Global Navigation Satellite System (GNSS) with an Inertial Measurement Unit (IMU), LiDAR, radar and/or optical camera can be considered as a potential candidate to meet reliable precision and integrity of position solution. In this paper, presenting ESA’s project, Integrity Monitoring and Prediction Concept for Autonomous Car Resilience and Safety (IMPACARS) by fusing GNSS with RTK-based positioning, IMU and LiDAR. The approach have been tested in various road-traffic environment scenarios, urban, semi-urban and GNSS signal denied environment. The results presented is based RTK-based positioning with static base station and moving station as potential another vehicle on road within the ad-hoc vehicular communication environment. The field trials continue to navigate through multipath and GNSS signal blockage regions, which are common in cities due to the high-rise buildings, creating a concrete jungle. IMU’s can provide corrections to the position, velocity and acceleration of the vehicle through the 6 degrees of freedom generated by the internal gyros and accelerometers and using a dead reckoning approach. The LiDAR can produce a full 360? scan of the environment, from these scans local features can be identified to provide relative positioning based on landmarks and detect objects in the vehicle’s path to prevent a collision. The system is maintained and monitored by the proposed innovative integrity model suitable for level 5 autonomous vehicles by fusing GNSS, IMU and LiDAR. Using this approach, a preliminary study was conducted, comparing the integrity performance of GNSS standalone positioning and sensor-fusion positioning within high multipath or GNSS signal blockage environments. The scenario cases are an open sky urban environment, narrow view due to high-rise buildings and tree cover where the dense foliage dilutes the GNSS signals. It is concluded that the positioning integrity of the vehicle after fusion is much more accurate and reliable, resulting in a suitable Protection Level (PL) for a level 5 autonomous vehicle.
Published in:	Proceedings of the 2020 International Technical Meeting of The Institute of Navigation January 21 - 24, 2020 Hyatt Regency Mission Bay San Diego, California
Pages:	538 - 556
Cite this article:	Tiwari, R., Stacey, T., Toran, Felix, "Integrity Monitoring and Prediction Concept and Prototype for Fully Autonomous Vehicle Resilience and Safety," Proceedings of the 2020 International Technical Meeting of The Institute of Navigation, San Diego, California, January 2020, pp. 538-556. https://doi.org/10.33012/2020.17209
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In